Download Monetary policy and its implementation

Ingimundur Fridriksson, Governor of the Central Bank of Iceland
Monetary policy and its implementation
Speech at Roundtable meeting, February 20, 2007
The interest rate decision on February 8, 2007
On Thursday, February 8 this year the Central Bank of Iceland
announced that the Board of Governors had decided to leave the
policy interest rate unchanged at 14.25%. A press release
announcing the decision stated that inflation had slowed down and
the short-term inflation outlook had improved beyond what was
expected in the Central Bank’s most recent forecast.1 However,
inflation was still significantly in excess of target. The current
interest rate might suffice to bring inflation to target within an
acceptable period of time, but significant macroeconomic
imbalances were still present that posed a risk of rising inflation. If
the inflation outlook took a turn for the worse, the Central Bank
would respond.
Monetary policy and macroeconomic developments
In the joint declaration by the Government of Iceland and the Central
Bank of Iceland on March 27, 2001, the Bank was assigned the main
objective of promoting price stability. This was defined as a rate of
inflation, measured as the twelve-month increase in the CPI, of as
close to 2½% as possible. Broadly speaking, central banks have only a
single instrument, i.e. their policy interest rate, and can therefore only
attain one macroeconomic goal in the long run. Internationally, a wide
consensus has been achieved between governments and economists
that price stability is the most appropriate main objective of monetary
policy.
The Icelandic economy has witnessed turbulent times since 2001 and
undergone major structural changes, particularly in the financial
sector. Iceland moved on to an inflation target under conditions of
strong macroeconomic imbalances, one result of which was a
depreciation of the króna by one-third over a period of eighteen
months until near the end of 2001. This development led to a shortlived inflationary episode which peaked in January 2002. Afterwards
inflation slowed down sharply and moved below the 2½% target
before the end of 2002. It remained below or close to the target into
the second half of 2004, but has been above target since then. In the
current episode, the twelve-month inflation rate peaked in August
2006 at 8.4%. Since then it has been decelerating and the twelve1
The press release is posted on the Central Bank’s website, www.sedlabanki.is
1
month increase in the CPI measured 7.4% in February 2007. However,
the rise in the core indices remained above 8%. Cuts in indirect
taxation will bring down headline inflation very rapidly over the next
few months, although the reduction in underlying inflation will be
slower.
The Central Bank has explained in detail the reasons for the inflation
target overshoot over the past two to three years. Suffice it to mention
here that the episode of overheating that now appears to be coming to
an end originated in large-scale investments [in the power and
aluminium sectors] in east Iceland and elsewhere. This was
compounded by unforeseen changes in the credit market when banks
began offering mortgage loans in competition with the state-owned
Housing Financing Fund [HFF], after the latter announced a
substantial increase in its loan-to-value ratio and mortgage amounts.
Another unexpected event was the issue of króna denominated
Eurobonds [glacier bonds], one side-effect of which was to delay the
transmission of the Central Bank’s policy rate hikes across the
domestic yield curve. The background to these transactions is a global
glut of savings which has led to exceptionally low international
interest rates, coinciding with unusually high interest rates in Iceland.
One consequence was to divert much of the monetary policy impact
into the exchange rate channel. Another more positive consequence
was a surge in turnover in the domestic financial markets, which
deepened them by bolstering price formation and enabled them to
handle much larger trading volume smoothly, without significant price
volatility.
Policy rate changes are only passed through in full to domestic
demand with a sizeable lag. In a small, open economy like Iceland, the
impact is initially reflected in the exchange rate. Other things being
equal, a rise in the policy rate causes a temporary appreciation of the
króna, which squeezes the export and traded goods sectors and also
channels demand out of the economy. The banks’ surprise entrance
into the domestic mortgage market made it difficult for the Central
Bank to respond effectively. This spurred a surge in household
borrowing and private consumption funded by credit. For an extended
period the Central Bank’s policy rate hikes primarily impacted the
exchange rate and had a more subdued effect on other domestic
interest rates, partly because of the glacier bond issues. Fierce
competition in the mortgage market between the banks and HFF also
undoubtedly delayed the force of the impact that the policy rate
increases should have had on mortgage lending rates, including those
of the HFF.
It should be added that changes in economic aggregates are more
difficult to measure in the national accounts during rapid growth
episodes, as shown by sizeable revisions to leading preliminary
2
estimates. When the Central Bank raised its policy rate in 2004 and
2005 to counter mounting inflationary pressures, it did so on the basis
of available data on macroeconomic conditions that later turned out to
give a misleading picture of real developments. Revised statistics
published last year showed that the economy expanded by much more
in 2004 and 2005 than could be read from data at hand when the
Central Bank hiked its policy rate during those two years. When the
real outcome was known, it was obvious that the Central Bank should
have raised its policy rate by more and faster than it did at the time.
Monetary policy faces the challenge that the best available data are
always imperfect, but especially so during periods of rapid change.
As I mentioned earlier, the Central Bank left its policy rate unchanged
on the last interest rate decision date, February 8. In December it
raised the rate to 14.25%. This was the eighteenth hike since May
2004; prior to that, the policy rate had been held at 5.3% for more than
a year. The Central Bank noted that its December hike seemed to come
as a surprise. The introduction to Monetary Bulletin, which was
published on November 2, stated that a reduction in the policy interest
rate was still not in sight, in spite of substantially brighter inflation
prospects. On the contrary, it was impossible to rule out a further
policy rate hike in order to attain the inflation target within an
acceptable timeframe, and the analysis presented then implied that
there were still grounds for tightening the monetary stance further.
However, the Board of Governors had decided to postpone such a
measure for the time being. The next interest rate decision, to be
announced on December 21, 2006, would be based on the analysis in
the November Monetary Bulletin and additional data that became
available by that time. The Chairman of the Board of Governors
underlined this message from Monetary Bulletin in a speech a few
days later when he described the decision to leave the policy rate
unchanged on November 2 as a deferral of an increase rather than the
end of the tightening of the stance.
When it was decided to raise the policy rate on December 21, it was
known that inflation had slowed down to considerably below the midyear forecast rate. Inflation was mainly brought down by lower energy
prices, the impact of the appreciation of the króna since the middle of
the year, less-than-expected wage drift and a slowdown in house price
inflation. Domestic demand in Q3 was broadly in line with the Central
Bank’s November forecast. It was also known that the current account
deficit was wider than expected in Q3 and would significantly
overshoot the forecast for the year as a whole. No indications of easing
labour market pressures could be seen. Unemployment was negligible
and there was a large influx of foreign labour. Despite lower inflation
than in the Bank’s July forecast, there was still no sign that the target
would be attained within two years if the effect of indirect tax cuts on
headline CPI in the first half of 2007 was ignored. Furthermore, the
3
wide current account deficit and tight labour market aroused
suspicions that pressures in the economy were underestimated, and
thereby inflationary pressures as well.
As mentioned earlier, policy rate hikes primarily impact the exchange
rate. The króna appreciated substantially until near the end of 2005. By
then it was considerably stronger than was compatible with external
balance in the economy, as the Central Bank had explained in detail,
among other things in its December Monetary Bulletin that year. So it
was not surprising that the króna depreciated. What was unexpected
was how soon and how fast the króna fell. It depreciated by 19%
against the exchange rate index over 2006 as a whole and all the
weakening was confined to the first half of the year. So far in 2007 the
króna has been roughly 15% weaker on average than at the beginning
of 2006. By comparison, the US dollar depreciated against the euro by
almost 11%. For most of the period since the middle of 2006, the
króna has fluctuated within a fairly modest range. Since the end of the
sharp depreciation in the first half of last year the króna can be
described as relatively stable and marginally above its long-term
average value. At present, the real effective exchange rate is slightly
stronger than on average over the last quarter of a century.
A wide current account deficit could potentially erode confidence in
the króna. The Central Bank has underlined the importance of
restoring macroeconomic balance by realigning domestic demand with
output capacity, and of establishing a sustainable external balance by
such means rather than through a sharp currency depreciation that
would fuel inflation and other instabilities. Inflation developments
depend heavily on the króna not slipping much from its level over the
past few months, at least for as long as pronounced macroeconomic
imbalances remain. Confidence in the exchange rate will be a crucial
factor over the coming months. It should also be pointed out that
exchange rate movements can be the result of external events or
changes beyond our control, such as a sudden shift in international
investors’ risk perceptions or global credit supply. The exchange rate
of the króna is closely linked to developments in capital markets
elsewhere in the world, especially while the current account deficit is
as large as at present.
Inflation and macroeconomic forecasting
The Central Bank bases its monetary policy formulation and
implementation on the inflation and macroeconomic forecast that it
prepares three times a year and publishes in Monetary Bulletin.
Forecasts are always uncertain and by definition can never state
precisely what will happen over the forecast horizon. The Central
Bank forecasts are based on extensive research and analysis, and on
models that have been developed to simulate how events are likely to
4
unfold. A major advance was made in 2006 with the introduction of a
new macroeconomic model designed by Central Bank economists. The
Central Bank has always expressed qualifications about the accuracy
of its forecasts, and readers of Monetary Bulletin will be familiar with
the risk profiles that draw attention to forecasting uncertainties.
Another uncertainty is the output gap at any time, which is an
important determinant of the inflation outlook. However, the economic
laws are known and are the same as everywhere. Forecasts are
prepared on the basis of the best available information and known
correlations between changes in aggregates. They therefore provide a
sufficiently good indication about future developments to enable the
Board of Governors to make decisions on the basis of them. Better
forecasts have not been available.
Nonetheless, the Central Bank has frequently been criticised for the
quality of its forecasting and the allegedly incorrect monetary policy
decisions based on them. For example, the Bank came under harsh fire
for raising the policy rate in the second half of 2006. It was claimed
that the forecasts were wrong and the hikes therefore unnecessary. In
some cases, critics selected limited and minor areas of Central Bank
forecasts which they considered to be based on incorrect assumptions,
and dismissed the entire forecast as a result.
Now that the year is over, it is obvious, of course, that the Bank’s
forecasts did not fully hold. For instance, in July 2006 the Bank
forecast higher second-half inflation than proved to be the case.
Explanations for the deviation include the appreciation of the króna
after the middle of the year, the effect of lower global fuel prices on
the price level, and the fact that wage settlements made in the middle
of the year did not provoke wage drift on the scale that the Bank had
forecast in July. The first two factors were hardly foreseeable, and it
should be pointed out that the Bank’s overforecast for wage drift was
not far from that made by the social partners themselves when the
wage settlements were signed. Be that as it may, the economy was no
less overheated than forecast in mid-2006, as witnessed by the current
account deficit and tight labour market. Thus it is untenable to claim
that the Central Bank’s policy rate hikes in 2006 were unwarranted.
On the contrary, a higher policy rate was the main precondition for the
króna to hold relatively strong, and contributed to subduing demand
growth. Had the Central Bank not raised its policy rate, there is little
doubt that inflation would be higher than at present.
Some commentators who were surprised by the December policy rate
hike apparently considered that the Central Bank had underestimated
the scope of the turnaround in the economy in the second half of the
year. National accounts statistics for Q3 that were published in midDecember, supposedly showing a substantial slowdown in economic
activity, were repeatedly cited as evidence. The Central Bank
considers it imprudent to draw sweeping conclusions solely from
5
preliminary estimates in the quarterly national accounts, which tend to
undergo sharp revisions. For this and other reasons, a broader range of
indicators contributed to the December policy rate decision, including
the current account deficit and labour market conditions.
The scenario today is that commercial bank analysts expect the cycle
of policy rate cuts to begin later this year than they were forecasting
towards the end of 2006. In other words, they consider that more
underlying pressures remain in the economy than they may have
thought at the time.
Interest rate assumptions in Central Bank forecasts
Various changes have been made to the exchange rate and interest rate
assumptions behind the Central Bank’s forecasts in recent years. In the
early years of inflation targeting, forecasts were based on the
assumption of an unchanged exchange rate and unchanged policy rate
over the forecast horizon. By showing how developments would
unfold if the Bank left the policy rate unchanged, they indicated
whether it needed to be changed. In December 2004, the first
alternative scenarios to the baseline forecast were presented which
assumed changes in the exchange rate and interest rate over the
forecast horizon. In December 2005 the same baseline forecast was
used, but in the alternative scenarios, the policy rate path followed
financial market analysts’ forecasts and the exchange rate was
extrapolated from uncovered interest rate parities. These approaches
were developed further in 2006 and the baseline forecast in July and
November was based on a policy rate path calculated from forward
interest rates and analysts’ expectations. The alternative scenarios
assumed, respectively, an unchanged policy rate and a policy rate path
that would bring inflation to target within the forecast horizon.
Earlier forecasts based on an unchanged policy rate across the horizon,
in an environment where there was an obvious need for a sharp hike,
produced a misleading picture. They presented developments that the
Bank had to explain would never materialise. The baseline forecast in
Monetary Bulletin in November 2006 represented an enhancement
insofar as it was based on market expectations about policy rate
changes. The Central Bank incorporated these expectations into its
baseline forecast, even though it considered them unrealistic. They
assumed rapid cuts in the policy rate commencing early this year. The
Central Bank considered neither assumption to be compatible with the
inflation outlook. Thus the forecast lacked credibility, in the Bank’s
own opinion.
This prompts the question of what assumptions it is natural for the
Central Bank to make about the policy rate in its baseline
macroeconomic and inflation forecasts.
6
Iceland is not the only country whose central bank has faced such a
challenge. Several years ago, the Reserve Bank of New Zealand
responded by using its own policy rate forecast as an assumption. The
Bank of Norway followed suit two years ago and now uses a future
interest rate path forecast by its Executive Board for its baseline
forecast. Sweden’s Riksbank recently announced that it will use its
own forecast for the repo rate path as an assumption in its inflation
forecast. These banks have followed a similar course to the Central
Bank of Iceland in developing the interest rate assumptions behind
their forecasts. They have now opted to use the path that is forecast by
those who actually determine the policy rate, partly because they did
not consider market agents’ forecasts and expectations sufficiently
credible. At the same time, these central banks have firmly stated that
their policy rate forecasts entail no commitment to hold their rates to
that path. The policy rate must always be decided on the basis of the
best available information at any given time. Experience shows that
paths change sharply from one forecast to the next and the central
banks explain the assumptions behind such changes thoroughly.
Above all, these central banks aim to exert a specific effect on
expectations, which is the principle behind a successful monetary
policy. This appears to have produced results in New Zealand and
Norway.
Another option is for the baseline forecast to assume a policy rate path
calculated from a monetary policy rule that steers inflation to target
within the forecast horizon. This was done in the alternative scenario
published in the November 2006 Monetary Bulletin.
Other inflation-targeting central banks take different approaches,
including the assumption of an unchanged policy rate as the Central
Bank of Iceland initially did after moving onto an inflation target.
The choice of assumptions for the Central Bank’s next macroeconomic
and inflation forecast has yet to be decided. It is still under
consideration and discussion in-house. What is crucial for the Central
Bank is the credibility of its forecasting and the ability to exert the
impact on expectations that it deems necessary.
Conclusion
It is not the Central Bank of Iceland’s policy to maintain high interest
rates. However, this is necessary for as long as the recent
macroeconomic conditions prevail. High interest rates subdue private
consumption and investment. The impact of interest rate changes is
transmitted with a lag, but monetary policy gradually delivers results,
now as ever. Reducing imbalances and restoring stability is in
everyone’s interest. Some strain is inevitable during an adjustment
from heavy imbalances, so complaints come as no surprise. But this
7
does not mean that the wrong monetary policy measures are being
applied – it could even mean quite the opposite.
The Central Bank’s policy rate is high. The press release announcing
the Board of Governors’ decision on February 8, which I quoted
earlier, stated that the current interest rate may suffice to bring
inflation to target within an acceptable period of time. It is impossible
to state this categorically now, because it depends on the inflation
outlook. What is certain is that the policy rate will be lowered as soon
as the Central Bank deems this compatible with the inflation target. It
is wrong to claim that this is ruled out because the interest rate
differential with abroad must be maintained to prevent turbulence
when glacier bonds mature. However, an untimely reduction in the
policy rate will benefit no one.
It has been claimed that the Central Bank is undermining the króna
with its interest policy, which will cause businesses and households to
abandon the currency. In this context it must be remembered that
without a tight monetary stance and high interest rates, the króna
would probably have depreciated by more than it actually did, demand
would have grown faster and inflation moved much higher. This
would have had serious consequences for indebted households and
businesses, as well as causing turmoil in the labour market and,
ultimately, even more instability in the economy.
Other central banks have periodically needed a very tight monetary
stance to bring inflation under control. For example after 1980,
roughly a quarter of a century ago, the US federal funds rate was
significantly higher for a while than Iceland’s current policy rate. This
proved necessary in order to rein in inflation that was also higher then
than in Iceland today. This policy was successful and the US has not
faced a serious challenge from inflation since.
Ideas for an economic panacea have also been aired, typically
involving the adoption of the euro. The Central Bank has already
stated its view that adopting the euro would be imprudent except by
joining the EU. Even though Iceland were to decide to apply for EU
membership, it would not be able to enter into the Monetary Union
until several years later, after fulfilling all the accompanying
conditions which include criteria for inflation, interest rates and
exchange rate stability. Experience shows that no concessions are
made from these conditions. It is also important to remember that
membership of a larger currency area does not eliminate the need for
sensible economic policy. What changes is that monetary policy
independence is lost. Decisions on consumer prices and unit labour
costs relative to abroad would be transferred to the respective markets.
This would make fiscal and wage flexibility even more vital in order to
tackle cyclical fluctuations. In the current adjustment of the economy,
the exchange rate of the króna has played a key role. Without a
8
flexible exchange rate, the crucial instruments for tackling prevailing
economic conditions would have been tight fiscal restraint, cuts in
public spending and/or higher taxes, to help ensure that sharp rises in
wages and prices did not excessively weaken the competitive position
of Iceland’s trading sectors.
Exchange rate volatility in the recent term was fuelled by the
overheating of the economy, which is now cooling once more. The
crucial factor for near-term exchange rate developments is to restore
economic stability in Iceland with inflation on target, and to secure it
in the longer run with prudent economic policies. It is doubtful
whether claims that economic policy would necessarily become more
prudent, if Iceland switched to the currency of a larger currency area,
are justifiable.
9